feat!: added hash index to pedersen

[sirasistant]

Related issue(s)

Related issue: noir-lang/noir#1341

Description

Adds the hash index constant parameter to pedersen bb func call.

Summary of changes

(Describe the changes in this PR. Point out breaking changes if any.)

Dependency additions / changes

(If applicable.)

Test additions / changes

(If applicable.)

Checklist

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt with default settings.
• I have linked this PR to the issue(s) that it resolves.
• I have reviewed the changes on GitHub, line by line.
• I have ensured all changes are covered in the description.

Additional context

(If applicable.)

[guipublic]

What this is doing?
Do we have a corresponding GitHub issue to motivate the change?

[TomAFrench]

Same, I'm missing a lot of context around this change so it would be helpful to have a full issue on what this is trying to solve.

[sirasistant]

Related issue: noir-lang/noir#1341

[guipublic]

This new field indicates which set of generators to use. The fact that we use it in Aztec for domain separation is not relevant in this context. A better name would be something like generators_set or generators_index.

[kevaundray]

This new field indicates which set of generators to use. The fact that we use it in Aztec for domain separation is not relevant in this context. A better name would be something like generators_set or generators_index.

Its main purpose is domain separation, so this name is fine

[kevaundray]

@guipublic

The naming of generator_set vs domain_separation is also related to noir-lang/noir#4760 ;

If we had a homomorphic pedersen implementation, then generator_set would be the correct choice, but since we are using it as a HashToGroup, domain_separation works better here.

[guipublic]

In any cases (homomorphic or not) it indicates which group elements to use, so generator_set is ok.
When the opcode will be renamed, then maybe another name will be better but we can (and should) change it at that time, not now.

[kevaundray]

In any cases (homomorphic or not) it indicates which group elements to use, so generator_set is ok.
When the opcode will be renamed, then maybe another name will be better but we can (and should) change it at that time, not now.

generator_set is not a detail that matters here, what we need is domain separation and knowing what the generator_set is, does not help in anyway due to this function not being homomorphic right now.

Lets go with domain_separatorand we can come back to it, if it proves to not be sufficient.

[guipublic]

generator_set is not a detail that matters here, what we need is domain separation

generator_set is a detail that does not matter for Aztec, what we need for Aztec is domain separation. When we will set this value coming from a noir contract, it will be called domain_index or domain_separator.

But at the acir level, domain_separator for a pedersen opcode is meaningless.

and knowing what the generator_set is, does not help in anyway due to this function not being homomorphic right now.

Knowing what the generators are allows someone to know what this opcode is doing, not what is his purpose, but what exact computation it is proving.

.